DRO Deakin Research Online, Deakin University’s Research Repository Deakin University CRICOS Provider Code: 00113B

Surface functionalization of carbon fibers with active screen plasma

Citation: Corujeira Gallo, Santiago, Charitidis, Constantinos and Dong, Hanshan 2017, Surface functionalization of carbon fibers with active screen plasma, Journal of vacuum science & technology A: vacuum, surfaces, and films, vol. 35, no. 2, pp. 021404-1- 021404-10.

©2017, American Vacuum Society

This article may be downloaded for personal use only. Any other use requires prior permission of the author and AIP Publishing. The following article appeared in Journal of vacuum science & technology A and may be found at http://dx.doi.org/10.1116/1.4974913

Downloaded from DRO: http://hdl.handle.net/10536/DRO/DU:30090826

Surface functionalization of carbon fibers with active screen plasma

Santiago Corujeira Galloa)

School of Metallurgy and Materials, University of Birmingham, Edgbaston, Birmingham B15 2TT,United Kingdom

Constantinos CharitidisSchool of Chemical Engineering, National Technical University of Athens, Iroon Polytechniou 9,Zografou 157 80, Greece

Hanshan DongSchool of Metallurgy and Materials, University of Birmingham, Edgbaston, Birmingham B15 2TT,United Kingdom

(Received 3 October 2016; accepted 11 January 2017; published 24 January 2017)

The active screen plasma technology was used to functionalize carbon fibers and vitreous carbon

disks. The plasma treatment conditions were mapped using optical emission spectroscopy and the

functionalized surfaces were analyzed using scanning electron microscopy and atomic force

microscopy, x-ray photoelectron spectroscopy, and contact angle measurements. A relationship was

found between the active species in the plasma and the functional groups attached to the carbon surfa-

ces, which provides valuable information for the optimization of the active screen plasma treatment.

Moreover, the surface analyses were repeated over a period of 28 days to study the aging of the func-

tionalized surfaces in air. The hydrophobic recovery was modeled using a surface restructuring theory

which revealed a mean lifetime of 3.4 days for the functional groups. VC 2017 American VacuumSociety. [http://dx.doi.org/10.1116/1.4974913]

I. INTRODUCTION

Carbon fibers (CFs) are exceptional reinforcements for

polymer matrix composites. The fibers exhibit high specific

strength and stiffness, thermal stability, and corrosion resis-

tance.1 CF composites allow significant weight reductions,

which make them attractive materials for aerospace, trans-

portation, and wind power applications, among others. The

properties of carbon fiber reinforced polymers (CFRP)

depend, to a large extent, on the interaction between the

strong CFs and the soft polymer matrix, most frequently an

epoxy resin. Therefore, several strategies have been explored

to modify the surface of CFs in order to improve the nature

of the interface,2,3 including:

(1) removing contaminants and weakly bonded layers from

the surface of the CFs;

(2) applying a binder or a coupling agent;

(3) increasing the surface roughness to promote mechanical

interlocking;

(4) attaching functional groups to increase the compatibility

with the polymer.

A large number of techniques have been reported to mod-

ify the surface of CFs, ranging from the wet chemical pro-

cesses4–6 to sophisticated RF plasmas,7 electron beam

irradiation,8 and multistep treatments.9 However, some of

these processes reduce the tensile strength of the fibers,

while others are complex, expensive, or inappropriate for

large scale implementation. Therefore, there is a need for a

simple and effective technique to functionalize CFs, and

nonthermal plasmas are attractive because of environmental

and technical reasons.10

The interaction between the active species in the plasma

and the carbon surface determines the type of functionaliza-

tion. With this regard, the surface structure of the CFs affects

the attachment of functional groups.11 Low-modulus fibers

have a less-ordered structure, and the dangling carbon bonds

offer preferential sites for functional groups to attach.12 On

the other hand, high-modulus fibers have a high degree of

alignment and the exposed graphitic basal planes are very

inert. In spite of this, plasma treatments are capable of func-

tionalizing graphitic surfaces by creating active sites through

mild ion etching.13

The gases used for plasma functionalization of CFs

include oxygen, argon, nitrogen, ammonia, and several

hydrocarbons.3 Ammonia (or the equivalent nitrogen-

hydrogen gas mixture) is particularly interesting because the

-NHx groups create strong bonds with the epoxy resins used

in CFRPs.3,10,13 At the same time, ammonia (or N2-H2) plas-

mas produce a mild ion etching, which creates active sites

while minimizing the surface damage. The excessive

increase in roughness and the creation of surface defects

which reduce the mechanical properties of the CFs are

important barriers for the uptake of surface treatments.11

Active screen (AS) is an advanced plasma technology

developed in the early 2000s.14 The experimental set up has

been described elsewhere,15 and the benefits of the AS tech-

nology are related with the exposure of the treated surfaces

to active species in the plasma, with minimum or no damage

by ion bombardment. In recent years, the AS technique was

used to functionalize polymers with N2-H2 plasmas,16 and a

preliminary study showed potential for the successful func-

tionalization of carbon surfaces, mainly carbon fibers.17

In this study, the AS plasma technology was used to func-

tionalize CFs with N2-H2 gas mixtures. The processinga)Electronic mail: corujeis@bham.ac.uk

021404-1 J. Vac. Sci. Technol. A 35(2), Mar/Apr 2017 0734-2101/2017/35(2)/021404/10/$30.00 VC 2017 American Vacuum Society 021404-1

conditions were optimized for the treatment of carbon surfa-

ces, and the discussion relates the active species identified in

the plasma with the surface chemistry measured on the func-

tionalized CFs and the resulting contact angle. Some obser-

vations about the aging of the functionalized surfaces in air

are also presented and discussed.

II. EXPERIMENT

A. Carbon fibers

T-700 carbon fibers were acquired from Torayca and the

physical and mechanical properties of the fibers are summa-

rized in Table I. A segment of the yarn, approximately 50 cm

long, was coiled inside a cellulose thimble, and the fibers

were cleaned with acetone for 24 h using a Soxhlet glass-

ware; the volume of the thimble was filled with freshly con-

densed acetone every 3 min.18 The fibers were subsequently

transferred to a clean glass beaker and dried inside an oven

for 7 h at 80 �C. The clean and dry fibers were stored in a

vacuum desiccator.

Short segments of CF were fixed onto holders made of

stainless steel foil (15 � 10� 0.1 mm) using double sided

copper tape for scanning electron microscopy (SEM). The

stainless steel foil holders had a 5 mm hole punched at the

center, where the fibers could be analyzed with minimum

background signal [Fig. 1(a)]. The holders also fitted tightly

inside glass vials, so that the fibers fixed onto them would

not touch the inner surfaces, thus avoiding contamination.19

Small pieces of aluminum foil were used as plugs, to limit

the movement of the foil along the vial.

B. Vitreous carbon disks

A vitreous carbon rod, 6 mm in diameter and 100 mm long,

was acquired from Almath Crucibles Ltd. (VCR6100). Disks

6 mm in diameter and 1 mm thick were extracted from the rod

using a Struers Accutom precision cut-off machine, fitted with

an Al2O3 blade and circulating abundant cooling fluid. The

disks were ground with wet abrasive papers #1200, #2500,

and #4000, and polished with colloidal silica suspension

(Struers OP-S). Finally, the polished disks were sonicated in

acetone for 10 min and dried with cold air [Fig. 1(a)].

C. Active screen plasma treatments

The plasma treatments were conducted inside a Klockner

Ionon 60 kW DC plasma furnace, where an AS arrangement

was prepared [Fig. 1(b)]. The cathodic mesh of the AS set-

up was made of austenitic stainless steel, with 6 mm holes,

disposed in a staggered arrangement, and a distance of

10 mm between centers. The vitreous carbon disks and the

pieces of stainless steel foil, with the carbon fibers attached

onto them, were placed on an insulated work table, which

was left at floating potential during the plasma treatment.

The distance between the samples and the lid of the AS set-

up was 15 mm. The temperature was monitored using a K-

type thermocouple inserted in a dummy block and remained

below 100 �C in all cases.

The vacuum chamber of the furnace was evacuated to a

base pressure below 1 Pa, at which point a gas mixture con-

sisting of 25% N2 and 75% H2 was fed into the chamber at a

flow rate of 4 l/h. The glow discharge was ignited between

the mesh of the AS set-up, which was connected to the cath-

ode, and the vacuum chamber of the furnace, which was

grounded. The gas pressure selected for the treatments was

50 Pa, the applied potential was between 200 and 400 V, the

current was limited to 1 A, and the treatment time ranged

from 5 to 40 min.

D. Optical emission spectroscopy

Optical emission spectra were acquired for different proc-

essing conditions using a Jobin Yvon Triax 180 monochroma-

tor, fitted with a 150 grooves/mm grating. The SPECTRAMAX for

Windows software was used to control the monochromator

and to collect the optical signal. The integration time was

30 s, and the slit width was 0.2 mm. The input optics consisted

of a fused silica optical fiber (UV-Vis multimode), with a core

diameter of 200 lm and a doped silica cladding 10 lm thick

TABLE I. Nominal properties of Toray carbon fibers T-700-12 k-60E (from spec sheet).

Strength (GPa) Modulus (GPa) Strain (%) Yield (g/1000 m) Density (g/cm3) Diameter (lm) Sizing type Sizing volume

4.9 230 2.1 800 1.80 7 60E 0.3%

FIG. 1. (Color online) (a) Macrograph of the specimens used for characterization and (b) active screen experimental set-up.

021404-2 Corujeira Gallo, Charitidis, and Dong: Surface functionalization of CFs 021404-2

J. Vac. Sci. Technol. A, Vol. 35, No. 2, Mar/Apr 2017

and a polyimide buffer (AMS Technologies SFH200/220/

245T). The fiber was fed into the vacuum chamber through a

sealed aluminum flange (custom made feedthrough for optical

fibers), and the tip was placed in close proximity to the sam-

ples, inside the AS experimental set-up. The data sets were

collected at cathodic potentials between 200 and 400 V and at

gas pressures between 25 and 125 Pa. The measurements

were acquired in duplicates, and two separate runs were con-

ducted, giving a total of four measurements per treatment con-

dition. The active species were identified using the PLASUS

SPECLINE v1.3 software.

E. Characterization

The surface morphology of the plasma treated fibers was

observed using a JEOL-7000 FEG SEM and a Veeco

MultiMode atomic force microscope (AFM), fitted with a

SiN cantilever for contact mode. The elastic constant of the

selected tip was 0.12 N/m.

The surface composition was assessed by x-ray photo-

electron spectroscopy (XPS) in a Thermo Scientific K-Alpha

instrument with a monochromatic source (1486.6 eV, 12 kV,

and 3 mA anode current). The spot size used was 400 lm,

and the pass energy values were 200 eV for the survey spec-

tra and 40 eV for the high-resolution scans, C (1s), N (1s),

and O (1s). Three data sets were collected for each specimen,

and all spectra were calibrated using the C (1s) sp2 reference

peak (284.5 eV) and subsequently added. The resulting spec-

tra were analyzed using CASA XPS version 2.3.16. The relative

sensitivity factors were 1.00 for C, 1.80 for N, and 2.93 for

O. The synthetic peak models were obtained using a

Tougaard background and Gaussian-Lorentzian peak shapes.

The components considered for the models are listed in

Table II. Several XPS measurements were conducted over a

period of 1 month, in order to determine the degradation or

aging of the functionalized surfaces in air.

F. Wettability

The contact angle measurements were obtained in an

Attension optical tensiometer, from Biolin Scientific, using

the sessile droplet method with deionized (DI) water. The

quantitative assessment of wettability on single filaments of

CF proved to be very challenging. Several attempts were

made to measure the contact angle of microdroplets on car-

bon fibers using different liquids: water, glycerol, ethylene

glycol, tetradecane, etc. These attempts failed because it was

difficult to form droplets of consistent size on single carbon

fibers and, in some cases, the liquid evaporated too quickly

so that the size and shape of the small droplets changed dur-

ing the measurement. Therefore, vitreous carbon disks were

used as proxies of CFs for the contact angle measurements.21

Even though the density of active sites on vitreous carbon

disks may differ from that on CFs, the relative trends

observed in both cases are expected to be very similar, par-

ticularly regarding the degradation or aging of the functional

groups in air.

To conduct these measurements, a small droplet (0.5 ll)

was suspended from the needle of the tensiometer and

slowly brought in contact with the surface of the vitreous

carbon disks. The volume of the droplet was subsequently

increased to 2 ll, and the needle was carefully withdrawn

before the images were collected. The contact angle was

measured automatically using the ONEATTENSION software

(v2.5), and the reported values are the average of at least

three separate measurements. The assessment includes sev-

eral tests conducted over a period of 1 month after the

plasma treatment, in order to determine the degradation or

aging of the functionalized surfaces in air.

III. RESULTS

A. Optical emission spectroscopy

The typical optical emission spectrum collected during

the AS plasma treatments is illustrated in Fig. 2. In all cases,

the emission spectra were dominated by the emission lines/

bands corresponding to Ha (656.3 nm) and N2þ (391.4 nm)

ions, followed by the N2* (313.6 nm), NH (337.0 nm), and

H2 excited species.22 The intensity of the peaks correspond-

ing to nitrogen containing species was normalized with

respect to the Ha peak for the purpose of comparison.

The results recorded for different AS plasma conditions

are summarized in Table III. In general, the emission inten-

sity was directly proportional to the applied voltage and

inversely proportional to the gas pressure. On the other hand,

the relative intensity of the N2* and NH excited species

showed the opposite trend. This is noteworthy because the

TABLE II. List of component peaks considered for the XPS model (Ref. 20).

C (1s) N (1s) O (1s)

Binding energy [eV] Component peak Binding energy [eV] Component peak Binding energy [eV] Component peak

284.4–284.6 C sp2 398.4–399.0 Pyridine 531.2–531.4 �C¼O

284.9–285.1 C sp3 399.4–399.8 Amines, amides 532.0–532.2 C¼O

284.8–285.0 C-COOH 400.1–400.7 Pyrroline, pyridone 532.7–532.9 -OH, C-O-C

285.4–285.6 C-N 401.1–401.7 Pyridinium 533.5–533.7 -OH, C-O

286.1–286.4 C-O-H, C�N 403.7–404.3 Shake up 534.8–535.2 Chemisorbed H2O

287.4–287.6 C¼O 405.5–405.7 NOx 536.2–536.6 Physisorbed H2O

288.3–288.9 COOH 538.2–538.6 Shake up

290.1–290.5 Physisorbed H2O

291.3–291.7 Shake up

021404-3 Corujeira Gallo, Charitidis, and Dong: Surface functionalization of CFs 021404-3

JVST A - Vacuum, Surfaces, and Films

surface functionalization is frequently attributed to such

excited molecules, and the emission intensity is a measure of

the abundance of the species in the plasma. Therefore, the

identification of suitable plasma conditions to form the

appropriate active species in the plasma is important for the

optimization of the process.

B. Surface chemistry

The typical survey spectra obtained by XPS on untreated

(UT) and AS plasma treated CFs are illustrated in Figs. 3(a)

and 3(b), respectively. The main resonance peaks are labeled

C (1s), N (1s) and O (1s). The peaks observed at higher bind-

ing energy were attributed to Auger effect, whereas the two

minor peaks at 100 eV and 150 eV, which appeared in all the

specimens without a clear trend, were attributed to Si con-

tamination. The quantified data are summarized in Table IV.

The results in Table IV reveal an increase in the nitrogen

content of the CFs after the AS plasma treatments. It is interest-

ing to note that, for a given treatment time (e.g., 20 min), the

nitrogen content shows an inverse correlation with the applied

voltage, which is in agreement with the abundance of N2* and

NH excited species measured by optical emission spectroscopy

(OES). Iron was also detected on the surface of the CFs plasma

treated at higher voltage or for longer times (20–40 min). This

is attributed to the sputtering of material from the mesh of the

AS set-up and its subsequent redeposition on the CFs. The iron

deposited on the CFs exhibited varying degree of oxidation

(FeO, Fe2O3, F3O4, and FeOOH), which accounts for the

higher oxygen content observed on these specimens.

The high-resolution spectra corresponding to the C (1s),

N (1s), and O (1s) regions of the UT and AS plasma treated

CFs are shown in Fig. 4. The C (1s) spectrum of the UT CF

exhibits a peak of C-O-R groups, which is attributed to resi-

dues of the binder agent (sizing) left on the surface after the

cleaning procedure; the amount of binder is small, judging

by the peak shape. The main changes in the C (1s) spectrum

after the AS plasma treatment are observed in the high-

energy tail, particularly related with an increase in the con-

tent of COOR functional groups.

The most noticeable change in the spectrum was observed

in the N (1s) region, where the intensity of the signal

increased after the AS plasma treatment, and the shape of

the peak changed considerably. The difference was attrib-

uted to the introduction of several N and N-H groups,20,23

mainly amine and pyridone. The O (1s) region of the scan

exhibited a broadening toward the low-energy tail of the

peak, which was attributed to the quinone (–C¼O) and ester

(C¼O) components. The samples treated at higher power

and longer times also exhibited oxygen peaks associated

with iron oxides.

The XPS analysis was repeated at regular time intervals

on one same specimen, and the results are illustrated in

FIG. 2. (Color online) Typical optical emission spectrum observed in the AS

plasma treatments.

TABLE III. Emission intensity of active species for different AS plasma processing conditions.

AS parameters Emission intensity (cts) Relative intensity (%)

Pressure [Pa] Potential [V] N2* (313.6 nm) NH (337.0 nm) N2

þ (391.4 nm) Ha (656.8 nm) N2*/Ha NH/Ha N2

þ/Ha

25 200 14499 16177 22751 24309 60 67 94

300 14925 16736 25647 26934 55 62 95

400 19094 25734 65535a 65535a 29 39 N/A

50 200 15262 17624 20618 21841 70 81 94

300 14925 16736 25647 26934 55 62 95

400 17385 22573 48770 50407 34 45 97

75 200 14468 15820 18618 20222 72 78 92

300 17037 22939 38754 41124 41 56 94

400 18454 27193 44871 47898 39 57 94

100 200 14977 17268 17382 18940 79 91 92

300 18276 27361 37756 41223 44 66 92

400 17942 26245 38050 41972 43 63 91

125 200 14375 16174 14938 16022 90 101 93

300 16619 24522 24190 26316 63 93 92

400 16559 24683 26292 28217 59 87 93

aThese measurements reached the saturation limit of the detector; however, the width of the peak at such intensity was only a few pixels, so the real intensity

was very close the saturation limit.

021404-4 Corujeira Gallo, Charitidis, and Dong: Surface functionalization of CFs 021404-4

J. Vac. Sci. Technol. A, Vol. 35, No. 2, Mar/Apr 2017

Fig. 5. Once again, the most evident changes were observed

in the N (1s) region of the spectrum. The synthetic peak

revealed a considerable reduction of the amine component

and a split between the pyridine and pyridone components,

which shows a trend toward the shape observed on the

untreated CFs. The chemical analysis conducted on the sur-

vey scans revealed a slow but consistent decrease in the

nitrogen content over a period of 28 days, from 5.8 to 5.1 at.

%. No significant change in the oxygen signal was detected.

Unfortunately, the first XPS measurement (t0) was con-

ducted approximately 2–3 weeks after the plasma treat-

ment, so the chemical composition immediately after the

treatment could not be assessed. Further discussion is given

in Sec. IV B.

C. Wettability

Contact angle measurements were conducted on vitreous

carbon disks, plasma treated alongside the CFs. Figure 6

shows the water droplets on carbon disks treated under dif-

ferent conditions. Increased wettability was obtained in all

cases, but the most significant changes were observed on

those specimens which were plasma treated at low potential

(<300 V) and for short times (<20 min). Moreover, the mac-

rographs in Fig. 4 were taken three days after the plasma

treatment, and subsequent measurements were conducted

within 1 h of the surface treatment revealed contact angles

as low as 10�.The evolution of the contact angle over time was signifi-

cant, and the changes are illustrated in Fig. 7. The untreated

carbon disks exhibited a high contact angle with deionized

water, which is characteristic of hydrophobic surfaces. After

the AS plasma treatment (300 V for 5 min), the contact angle

dropped dramatically to values as low as 10�. However, the

samples exhibited a hydrophobic recovery over time, as the

plasma treated surface was exposed to atmospheric condi-

tions. The progression was very rapid in the first few hours

or days after the plasma treatment, but it became slower after

1 week and the contact angle finally stabilized at approxi-

mately 70�, which is slightly lower than the untreated

material.

It is important to mention that the measurements were

conducted both on pristine samples and on one same speci-

men at several aging times. For a given aging time, the

specimens used for multiple measurements exhibited mar-

ginally higher contact angles than the pristine specimens,

which had contact with the water for the very first time; but

the difference was within the experimental error. Therefore,

it was concluded that the exposure to deionized water during

the measurements only had a minor effect on the results, and

the aging was mainly attributed to the exposure of the AS

plasma treated surfaces to air.

D. Surface morphology

The surface morphology of the untreated and AS plasma

treated CFs are illustrated in Fig. 8. A smooth and featureless

surface was observed in all cases, and the changes introduced

by the AS plasma treatment were very minor, even under the

most severe treatment conditions. The absence of ion bom-

bardment on the CFs is evident from the lack of surface

defects or arcing damage. No obvious signs of intense chemi-

cal etching were observed either, and only the longitudinal

grooves of the CFs appeared somewhat deeper after the plasma

treatment. Therefore, the small changes in surface roughness

were attributed to the mild chemical etching and the deposition

of iron particles sputtered from the AS set-up. No changes in

the surface morphology were observed upon aging in air.

IV. DISCUSSION

A. Active species responsible for the surfacefunctionalization

A comparison of the OES and XPS data presented in

Sec. III reveals a correlation between the excited species in

FIG. 3. (Color online) X-ray photoelectron spectra corresponding to (a) untreated and (b) AS plasma treated carbon fiber (300 V, 5 min).

TABLE IV. Summary of XPS results obtained from carbon fibers (at. %).

Condition C (1s) N (1s) O (1s) Fe (2p)

UT 84.82 6 3.69 1.92 6 0.15 13.66 6 3.59 N/A

AS 200 V 20 min 80.36 6 1.52 5.59 6 0.24 13.87 6 1.56 0.18 6 0.02

AS 300 V 5 min 82.19 6 2.17 4.90 6 0.06 12.88 6 2.19 0.03 6 0.01

AS 300 V 20 min 79.52 6 1.93 5.08 6 0.24 15.21 6 2.03 0.19 6 0.02

AS 300 V 40 min 76.63 6 0.69 5.88 6 0.22 15.73 6 0.14 1.76 6 0.41

AS 400 V 20 min 67.78 6 3.14 4.23 6 0.17 23.40 6 2.97 4.58 6 0.35

021404-5 Corujeira Gallo, Charitidis, and Dong: Surface functionalization of CFs 021404-5

JVST A - Vacuum, Surfaces, and Films

FIG. 4. (Color online): High resolution spectra for (a) and (b) C (1s); (c) and (d) N (1s); and (e) and (f) O (1s) corresponding to untreated (UT) and AS plasma

treated carbon fibers.

FIG. 5. (Color online) Evolution of the N (1s) peak over time for sample AS 300 V 20 min: (a) first measurement (t0) and (b) fourth measurement (t0 þ 28

days).

021404-6 Corujeira Gallo, Charitidis, and Dong: Surface functionalization of CFs 021404-6

J. Vac. Sci. Technol. A, Vol. 35, No. 2, Mar/Apr 2017

the plasma, particularly NH and N*, and the nitrogen content

on the surface of the plasma treated CFs (Fig. 9). It is inter-

esting to note that the ionized species in the plasma, mainly

N2þ, showed no correlation with the nitrogen content on the

CFs. Therefore, the functionalization of the carbon fibers

may be linked to the excited species in the plasma, NH and

N*, rather than the ionized N2þ species.

Unfortunately, plasma diagnostic techniques have not

been extensively used to characterize the plasma functionali-

zation of CFs. However, several studies have been published

on the surface functionalization of polymers with N2-H2

plasmas, and some observations may be extrapolated to car-

bon surfaces.24 With regards to the attachment of nitrogen

functionalities to the treated surface, particularly amine

groups, other authors have reported a dependence on the

applied power, gas mixture, and gas pressure.24,25 The

observed trend is similar to the one described in this study,

with higher yield of nitrogen functionalities at low power

and higher pressure.26 The mechanism behind this correla-

tion with the applied power has been attributed to changes in

the plasma chemistry, mainly dissociation of N2, H2, and

NH2/NH3 species at higher power.24,26

In addition to this, Favia et al.24 suggested that there was

a direct correlation between nitrogen functionalities grafted

on the treated surface, and the N2* and NH active species in

the plasma. This, however, does not exclude the role of other

species, which may not be detected by OES. A high content

of hydrogen in the gas mixture is also said to create active

sites through a selective etching mechanism which facilitates

the attachment of functional groups. In the case of carbon

surfaces, a similar etching mechanism was reported by Yang

et al.27 for graphene exposed to a hydrogen RF plasma. The

mild ion bombardment by positive ions was also said to cre-

ate active sites for the attachment of functional groups.12

However, this mechanism should be negligible in the AS

arrangement, given that the CFs are kept at floating potential,

and without cathodic bias.

On the other hand, the ionized N2þ species showed a

direct correlation with the iron content measured on the

fibers after the AS plasma treatment. This is attributed to the

sputtering effect of energetic ions on the AS set-up, and the

deposition of the sputtered material onto the exposed surfa-

ces, including the CFs. The effect of such deposition layer

on the CFs, and on the bond with the polymeric matrix in

composite materials, has not been assessed yet. However,

curing kinetic studies of epoxy resin modified with iron and

iron oxide (Fe2O3) nanoparticles suggest that the iron depo-

sition layer is likely to result in higher cross-linking density,

glass transition temperature, thermal stability, and improved

mechanical properties.28,29

B. Aging of functional groups attached to the surfaceof CFs

With regards to the contact angle measurements, Narushima

et al.30 reported that low power plasma treatments resulted

in lower contact angles, thus following a similar trend to the

functional groups attached on the surface. In the present

study, the change in contact angle observed after the AS

plasma treatment was remarkable. However, a hydrophobic

recovery took place upon exposure of the specimens to air

[Fig. 10(a)]. This behavior was observed in other materials,

mainly polymers, although the rate of decay depended on the

nature of the substrate, the plasma treatment conditions, and

the aging environment.31–34 In general, the change in chem-

istry and properties of the plasma functionalized surfaces is

attributed to two factors, namely, surface oxidation upon

reaction with atmospheric oxygen and surface adaptation

through diffusion of surface species in the bulk,35,36 although

the latter mechanism is assumed to be negligible for carbon

substrates.

The model proposed by Israelachvili and Gee37 can be

used to calculate the fraction of polar and nonpolar regions

on plasma functionalized surfaces38,39

FIG. 6. Sessile droplets on vitreous carbon disks after AS plasma treatments under different conditions.

FIG. 7. Evolution of the contact angle after the AS plasma treatment at 300 V for 5 min. The samples were subsequently aged in air.

021404-7 Corujeira Gallo, Charitidis, and Dong: Surface functionalization of CFs 021404-7

JVST A - Vacuum, Surfaces, and Films

ð1þ Cos hÞ2 ¼ fp ð1þ Cos hpÞ2 þ fnp ð1þ Cos hnpÞ2;(1)

where h is the equilibrium contact angle measured on the

plasma treated surface, fp is the fraction of the surface cov-

ered by polar regions, fnp is the fraction of the surface cov-

ered by nonpolar regions (fp¼ 1 – fnp), hp is the equilibrium

contact angle on a surface consisting of polar regions only

(assumed to be 0�), and hnp is the equilibrium contact angle

on a surface consisting of nonpolar regions only (assumed to

be 85�). Figure 10(b) shows the evolution of fp with aging

time. The initial value of fp immediately after the plasma

treatment is 97.5%, which correlates with the expected

attachment of functional groups with N2-H2 plasma treat-

ments. The lowest fp value calculated after 28 days of aging

in air was 17%, which illustrates the hydrophobic recovery

described before. The change observed upon aging in air had

an exponential decay form, described by

fp ¼ y0 þ A1 � e�ðx=sÞ: (2)

The fitting parameters and the quality of the fitting are

reported in Table V. The fact that the experimental data can

be modeled with an exponential decay function of the first

order supports the previous assumption that the aging pro-

cess may be attributed to one single physical mechanism,

namely, the exposure of the plasma treated surface to air,

while the diffusion of surface functionalities into the bulk

material is negligible in this case. The value of s in Eq. (2)

represents the mean lifetime of the functional groups on the

carbon surface, i.e., 3.4 days, and yo is the residual value of

fp after aging, which in this case was 22%.

The AS plasma treated CFs exhibited a very similar sur-

face morphology to the untreated fibers. The small differ-

ences were attributed to the mild chemical etching of

hydrogen as well as the deposition of iron from the AS set-

up. In any case, no changes in surface morphology were

observed over time. Therefore, there is no evidence to sup-

port that the changes in contact angle are associated with the

topography of the carbon surfaces, and thus, the changes are

mainly attributed to the surface chemistry.

The AS plasma treatments with a N2-H2 gas mixture

clearly introduced polar functional groups on the carbon sur-

faces. This was reflected on the chemical composition mea-

sured by XPS and on the contact angle with DI water.

However, the functional groups underwent an aging process

when exposed to air. The contact angle could be measured

within 1 h of the plasma treatment and at frequent time inter-

vals over a period of 1 month. Therefore, this data set

reflects the initial condition of the surface and its evolution

over time with significant fidelity.

On the other hand, the first XPS measurement was con-

ducted 2–3 weeks after the plasma treatment, when the func-

tional groups had already degraded to some extent. As aFIG. 9. (Color online) Correlation between the excited species in the plasma

(OES) and the resulting nitrogen content on the carbon fibers (XPS).

FIG. 8. (Color online) SEM micrographs and AFM surface plots of (a) and (c) untreated CF, (b) and (d) AS plasma treated CF (300 V, 20 min).

021404-8 Corujeira Gallo, Charitidis, and Dong: Surface functionalization of CFs 021404-8

J. Vac. Sci. Technol. A, Vol. 35, No. 2, Mar/Apr 2017

consequence, the drop in nitrogen content observed in the

XPS data set is attributed to the tail of the exponential decay

described by Eq. (2). The extrapolation of the nitrogen con-

tent to the initial value using the lifetime (s) and pre-

exponential factor (A1) in Table V revealed values as high as

80% immediately after the plasma treatment and 15% after 1

week.

The XPS results showed a decrease in the nitrogen con-

tent over time, but the oxygen content remained constant.

Unfortunately, the degradation sequence is difficult to iden-

tify by XPS surface analysis because of technique limita-

tions.40 It is speculated that the NH functional groups

undergo oxidative scissions and evaporative loss upon their

exposure to air, although the elucidation of the degradation

mechanism requires further study.

V. SUMMARY AND CONCLUSIONS

The functionalization of carbon fibers with the active

screen plasma technology was successful. The surface of

the CFs was enriched with nitrogen functional groups,

which could be linked to NH and N* excited species in the

plasma. Furthermore, the map of processing conditions pro-

duced by OES provides general guidelines for future pro-

cess optimization. With this regard, active screen plasma

conditions combining low voltage and relatively high pres-

sure are favorable for the formation of the excited species

mentioned previously.

The active screen plasma treatment also produced a

remarkable reduction in the contact angle of the carbon sur-

faces. However, the surfaces exhibited a hydrophobic recov-

ery upon their exposure to air. The exponential decay, based

on a surface restructuring model, revealed that the mean life-

time of the functional groups on the carbon surface is 3.4

days, and the residual fraction of polar regions after aging is

22%. The aging mechanisms remain unclear, but they seem

to involve oxidative scissions and evaporative loss of nitro-

gen functional groups.

ACKNOWLEDGMENTS

The research leading to these results has received funding

from the European Union’s Seventh Framework Program

(FP7/2007-2013) under the Agreement No. GA604248. X-

ray photoelectron spectra were obtained at the National

EPSRC XPS Users’ Service (NEXUS) at Newcastle

University, an EPSRC Mid-Range Facility.

1M. Sharma, S. Gao, E. M€ader, H. Sharma, L. Y. Wei, and J. Bijwe,

Compos. Sci. Technol. 102, 35 (2014).2S. Tiwari and J. Bijwe, Procedia Technol. 14, 505 (2014).3B. Z. Jang, Compos. Sci. Technol. 44, 333 (1992).4C. U. Pittman, Jr., G. R. He, B. Wu, and S. D. Gardner, Carbon 35, 317

(1997).5F. Severini, L. Formaro, M. Pegoraro, and L. Posca, Carbon 40, 735

(2002).6L. H. Meng, Z. W. Chen, X. L. Song, Y. X. Liang, Y. D. Huang, and Z. X.

Jiang, J. Appl. Polym. Sci. 113, 3436 (2009).7D. Akbar and T. E. G€ung€or, Surf. Coat. Technol. 240, 233 (2014).8L. Mao, Y. Wang, Z. Zang, S. Zhu, H. Zhang, and H. Zhou, J. Appl.

Polym. Sci. 131, 40274 (2014).9Z. Chen, X. J. Dai, P. R. Lamb, J. du Plessis, D. R. de Celis Leal, K.

Magniez, B. L. Fox, and X. Wang, Plasma Processes Polym. 10, 1100

(2013).10R. Morent, N. De Geyter, J. Verschuren, K. De Clerck, P. Kiekens, and C.

Leys, Surf. Coat. Technol. 202, 3427 (2008).11L. G. Tang and J. L. Kardos, Polym. Compos. 18, 100 (1997).12C. Jones and E. Sammann, Carbon 28, 515 (1990).13C. Jones and E. Sammann, Carbon 28, 509 (1990).14J. Georges, U.S. patent 5989363 (23 November 1999).15C. X. Li, T. Bell, and H. Dong, Surf. Eng. 18, 174 (2002).16X. Fu, M. J. Jenkins, G. Sun, I. Bertoti, and H. Dong, Surf. Coat. Technol.

206, 4799 (2012).17S. Corujeira Gallo, M. Koklioti, P. Jagdale, E. P. Koumoulos, H. Dong, A.

Tagliaferro, and C. A. Charitidis, paper presented at the Carbon 2015,

Dresden, 2015 (unpublished).18ISO-BSI, 10548: Carbon Fibre—Determination of Sizing Content (ISO,

London, 2003).19ISO-BSI, 18117: Surface Chemical Analysis—Handling of Specimens

Prior to Analysis (ISO, London, 2009).20F. Vautard, S. Ozcan, F. Paulauskas, J. E. Spruiell, H. Meyer, and M. J.

Lance, Appl. Surf. Sci. 261, 473 (2012).21J. B. Donnet and G. Guilpain, Carbon 27, 749 (1989).22R. W. B. Pearse and A. G. Gaydon, The Identification of Molecular

Spectra, 3rd ed. (Chapman & Hall Ltd, London, 1963).23J. R. Pels, F. Kapteijn, J. A. Moulijn, Q. Zhu, and K. M. Thomas, Carbon

33, 1641 (1995).24P. Favia, M. V. Stendardo, and R. D’Agostino, Plasmas Polym. 1, 91

(1996).25K. Narushima, M. Fukuoka, H. Kawai, N. Inagaki, Y. Isono, and M. R.

Islam, Jpn. J. Appl. Phys., Part 1 46, 7855 (2007).

FIG. 10. (Color online) Evolution with aging time in air: (a) contact angle with DI water and (b) calculated fraction of polar regions.

TABLE V. Fitting parameters for the exponential decay of fp.

yo A1 s R2

21.799 6 2.809 76.712 6 5.088 3.376 6 0.533 0.983

021404-9 Corujeira Gallo, Charitidis, and Dong: Surface functionalization of CFs 021404-9

JVST A - Vacuum, Surfaces, and Films

26A. A. Meyer-Plath, K. Schr€oder, B. Finke, and A. Ohl, Vacuum 71, 391

(2003).27R. Yang, L. Zhang, Y. Wang, Z. Shi, D. Shi, H. Gao, E. Wang, and G.

Zhang, Adv. Mater. 22, 4014 (2010).28O. Zabihi, M. Aghaie, and K. Zare, J. Therm. Anal. Calorim. 111, 703

(2013).29T. Sun, H. Fan, Z. Wang, X. Liu, and Z. Wu, Mater. Des. 87, 10

(2015).30K. Narushima, N. Yamashita, Y. Isono, M. R. Islam, and M. Takeuchi,

Jpn. J. Appl. Phys., Part 1 47, 3603 (2008).31W. J. Brennan, W. J. Feast, H. S. Munro, and S. A. Walker, Polymer 32,

1527 (1991).32A. Jord�a-Vilaplana, L. S�anchez-N�acher, D. Garc�ıa-Sanoguera, A.

Carbonell, and J. M. Ferri, J. Appl. Polym. Sci. 133, 43040 (2016).

33S. H. Lee, W. C. Lin, C. H. Kuo, M. Karakachian, Y. C. Lin, B. Y. Yu,

and J. J. Shyue, J. Phys. Chem. C 114, 10512 (2010).34A. Merenda, E. D. Ligneris, K. Sears, T. Chaffraix, K. Magniez, D. Cornu,

J. A. Sch€utz, and L. F. Dum�ee, Sci. Rep. 6, 31565 (2016).35K. S. Siow, L. Britcher, S. Kumar, and H. J. Griesser, Plasma Processes

Polym. 3, 392 (2006).36D. Hegemann, Thin Solid Films 581, 2 (2014).37J. N. Israelachvili and M. L. Gee, Langmuir 5, 288 (1989).38T. R. Gengenbach, Z. R. Vasic, S. Li, R. C. Chatelier, and H. J. Griesser,

Plasmas Polym. 2, 91 (1997).39R. C. Chatelier, X. Xie, T. R. Gengenbach, and H. J. Griesser, Langmuir

11, 2576 (1995).40A. Ahmadalinezhad, R. Tailor, and A. Sayari, Chem. Eur. J. 19, 10543

(2013).

021404-10 Corujeira Gallo, Charitidis, and Dong: Surface functionalization of CFs 021404-10

J. Vac. Sci. Technol. A, Vol. 35, No. 2, Mar/Apr 2017